亚纳秒钪锑碲缓存器件结晶动力学及纳米尺寸效应研究

The huge speed gap between the DRAM and NAND Flash has become the bottleneck of improving the overall computing ability in current computing hierarchy. Due to their non-volatility, high switching speed, low power consumption and high endurance, phase-change memory is deemed as the most promising candidate to bridge this performance gap. However, the crystallization speed of traditional phase-change materials (Ge2Sb2Te5) is still 10 times slower than that of DRAM. A recently developed phase-change material—Sc doped antimony telluride can crystallize in sub-nanosecond. Nevertheless, no conclusion has been made on the optimal composition and its nano-size effect. This proposal will focus on ScxSb2Te3 phase-change materials and investigate their (1) temperature-dependent crystallization kinetics, including viscosity, fragility, crystal growth rate and nucleation rate as a function of Sc contents via characterizing their heat flow curves at wide span of heating rates in association with subsequent crystallization modeling; (2) size effect of electrical performance and thermal stability by fabricating nanosized phase-change memory using electron beam lithography, in order to unravel the mechanism of Sc doping induced crystallization acceleration, obtain the optimal Sc content in ScxSb2Te3 alloy and achieve faster speed and lower power consumption in phase-change memory. This research can provide deep insights and theoretical guidance to the engineering of ScxSb2Te3 based phase-change memory.

目前存储器行业DRAM和NAND Flash之间巨大的速度差异是其计算性能提升的主要瓶颈。由于非易失性、高运算速度、低功耗、高循环寿命等优点，相变存储器被认为是最有希望填补此性能鸿沟的存储技术。然而传统相变材料（Ge2Sb2Te5）的相变速度和DRAM比仍然有不小的差距。最近开发出的一种新型相变材料钪锑碲具有亚纳秒相变速度，然而其体系的最优组分及纳米尺寸效应尚无定论。本项目聚焦能潜在应用于高速相变存储器的钪锑碲相变材料，拟主要通过（1）表征相变材料在宽加热速率下的热流曲线，建立结晶模型，结合数据分析研究其结晶动力学，如黏度、脆性、晶体生长速率、形核率等与温度及钪组分的对应关系；（2）采用电子束曝光设备制备纳米尺度相变存储器研究钪碲锑相变器件电学性能、热稳定性等的纳米尺寸效应，以阐明钪锑碲相变材料中钪掺杂加速材料结晶的作用机制、优选钪锑碲材料体系中钪组分；同时微缩化相变器件以加快结晶速度、降低功耗，为实现钪锑碲相变材料在自主闪存型相变存储器中的工程化提供理论基础。

作为新兴存储技术中最有潜力衔接传统计算架构中的DRAM+FLASH模式的存储技术，相变存储器在存算一体、类脑计算等新领域有着巨大的应用前景。最近新开发的一种新型相变存储材料，钪掺杂锑碲，将传统相变存储材料的结晶速度提高了一个数量级，但在组分优化、尺寸效应等方面缺乏系统研究，阻碍了其进一步应用。本项目聚焦钪锑碲相变材料的结晶动力学和器件性能：（1）通过研究钪锑碲相变材料在宽升温速率范围内的热流曲线，建立结晶模型研究其过冷液相的黏度、脆性、晶体生长速率、形核率等重要参数在玻璃化转变温度和熔点温区的变化，发现黏度的强-脆转变和异质形核促使钪碲锑能同时具有结晶速度快、非晶热稳定性好的优点。（2）采用微纳加工制备不同钪组分的纳米尺度相变存储器、系统研究钪碲锑相变器件电学性能、热稳定性等，结合结晶行为优选钪锑碲材料体系中钪组分,Sc组分为0.2-0.3。（3）研究微缩化相变器件性能，以解析其纳米尺寸效应，获得稳定的多态，相较于传统器件钪碲锑相变存储器漂移系数降低了约2-3数量级，为实现钪锑碲相变材料在自主闪存型相变存储器中的工程化提供理论基础。